Download presentation
Presentation is loading. Please wait.
1
Combinational Circuit Design
COE 202 Digital Logic Design Dr. Muhamed Mudawar King Fahd University of Petroleum and Minerals
2
Presentation Outline How to Design a Combinational Circuit
Designing a BCD to Excess-3 Code Converter Designing a BCD to 7-Segment Decoder Hierarchical Design Iterative Design
3
Combinational Circuit
A combinational circuit is a block of logic gates having: π inputs: π₯1, π₯2, β¦, π₯π π outputs: π1, π2, β¦, ππ Each output is a function of the input variables Each output is determined from present combination of inputs Combination circuit performs operation specified by logic gates οΌ Combinational Circuit π inputs π outputs
4
How to Design a Combinational Circuit
Specification Specify the inputs, outputs, and what the circuit should do Formulation Convert the specification into truth tables or logic expressions for outputs Logic Minimization Minimize the output functions using K-map or Boolean algebra Technology Mapping Draw a logic diagram using ANDs, ORs, and inverters Map the logic diagram into the selected technology Considerations: cost, delays, fan-in, fan-out Verification Verify the correctness of the design, either manually or using simulation
5
Designing a BCD to Excess-3 Code Converter
Specification Convert BCD code to Excess-3 code Input: BCD code for decimal digits 0 to 9 Output: Excess-3 code for digits 0 to 9 Formulation Done easily with a truth table BCD input: π, π, π, π Excess-3 output: π€, π₯, π¦, π§ Output is don't care for 1010 to 1111 BCD a b c d Excess-3 w x y z 1010 to 1111 X X X X
6
Designing a BCD to Excess-3 Code Converter
Logic Minimization using K-maps 00 01 11 10 ππ ππ K-map for π€ K-map for π₯ K-map for π¦ K-map for π§ 1 X Minimal Sum-of-Product expressions: π€=π+ππ+ππ , π₯=πβ²π+πβ²π+ππβ²πβ² , π¦=ππ+πβ²πβ² , π§=πβ² Additional 3-Level Optimizations: extract common term (π+π) π€=π+π(π+π) , π₯= π β² π+π +π π+π β² , π¦=ππ+(π+π)β²
7
Designing a BCD to Excess-3 Code Converter
Technology Mapping Draw a logic diagram using ANDs, ORs, and inverters Other gates can be used, such as NAND, NOR, and XOR Using XOR gates π₯= π β² π+π +π π+π β² =π ο
π+π π¦=ππ+ π β² π β² = π ο
π β² =π ο
πβ² a b c d w x y z a b c d w x y z
8
Designing a BCD to Excess-3 Code Converter
Verification Can be done manually Extract output functions from circuit diagram Find the truth table of the circuit diagram Match it against the specification truth table Verification process can be automated Using a simulator for complex designs Truth Table of the Circuit Diagram BCD a b c d c+d b(c+d) Excess-3 w x y z 1 a b c d w = a + b(c + d) x = b ο
(c + d) y = c ο
d' z = d'
9
BCD to 7-Segment Decoder
Seven-Segment Display: Made of Seven segments: light-emitting diodes (LED) Found in electronic devices: such as clocks, calculators, etc. BCD to 7-Segment Decoder A B C D a b c d e f g BCD to 7-Segment Decoder Accepts as input a BCD decimal digit (0 to 9) Generates output to the seven LED segments to display the BCD digit Each segment can be turned on or off separately
10
Designing a BCD to 7-Segment Decoder
Specification: Input: 4-bit BCD (A, B, C, D) Output: 7-bit (a, b, c, d, e, f, g) Display should be OFF for Non-BCD input codes Formulation Done with a truth table Output is zero for 1010 to 1111 Truth Table BCD input A B C D 7-Segment decoder a b c d e f g 1010 to 1111
11
Designing a BCD to 7-Segment Decoder
Logic Minimization Using K-Maps 1 00 01 11 10 π΄π΅ πΆπ· K-map for π K-map for π K-map for π π= π΄ β² πΆ+ π΄ β² π΅π·+π΄ π΅ β² πΆ β² + π΅ β² πΆ β² π· β² π= π΄ β² π΅ β² + π΅ β² πΆ β² + π΄ β² πΆ β² π· β² + π΄ β² πΆπ· π= π΄ β² π΅+ π΅ β² πΆ β² + π΄ β² π· Optimized Logic Expressions π= π΄ β² πΆ+ π 1 π·+ π 2 π΄+ π 2 π· β² π= π΄ β² π΅ β² + π 2 + π΄ β² πΆ β² π· β² + π 3 πΆ π= π 1 + π 2 + π 3 π 1 , π 2 , π 3 are shared gates Extracting common terms Let π 1 = π΄ β² π΅, π 2 = π΅ β² πΆ β² , π 3 = π΄ β² π·
12
Designing a BCD to 7-Segment Decoder
Logic Minimization Using K-Maps 00 01 11 10 π΄π΅ πΆπ· K-map for π 1 K-map for π K-map for π K-map for π Common AND Terms ο¨ Shared Gates π 4 =π΄ π΅ β² πΆ β² , π 5 = π΅ β² πΆ β² π· β² π 6 = π΄ β² π΅ β² πΆ, π 7 = π΄ β² πΆ π· β² π 8 = π΄ β² π΅ πΆ β² , π 9 = π΄ β² π΅π·β² Optimized Logic Expressions π= π 4 + π 5 + π 6 + π 7 + π 8 π· π= π 5 + π 7 π= π 4 + π 5 + π 8 + π 9 π= π 4 + π 6 + π 8 + π 9
13
Designing a BCD to 7-Segment Decoder
Technology Mapping T4 T2 T5 A B' C' D' T0 T6 T7 A' C T8 T1 T9 B e f g Many Common AND terms: π 0 thru π 9 π 0 = π΄ β² πΆ, π 1 = π΄ β² π΅, π 2 = π΅ β² πΆ β² π 3 = π΄ β² π·, π 4 =π΄ π΅ β² πΆ β² , π 5 = π΅ β² πΆ β² π· β² π 6 = π΄ β² π΅ β² πΆ, π 7 = π΄ β² πΆ π· β² π 8 = π΄ β² π΅ πΆ β² , π 9 = π΄ β² π΅ π· β² Optimized Logic Expressions π= π 0 + π 1 π·+ π 4 + π 5 π= π΄ β² π΅ β² + π 2 + π΄ β² πΆ β² π· β² + π 3 πΆ π= π 1 + π 2 + π 3 π= π 4 + π 5 + π 6 + π 7 + π 8 π· π= π 5 + π 7 π= π 4 + π 5 + π 8 + π 9 π= π 4 + π 6 + π 8 + π 9 Showing only Outputs e, f, g
14
Verification Methods Manual Logic Analysis Simulation
Find the logic expressions and truth table of the final circuit Compare the final circuit truth table against the specified truth table Compare the circuit output expressions against the specified expressions Tedious for large designs + Human Errors Simulation Simulate the final circuit, possibly written in HDL (such as Verilog) Write a test bench that automates the verification process Generate test cases for ALL possible inputs (exhaustive testing) Verify the output correctness for ALL input test cases Exhaustive testing can be very time consuming for many inputs
15
Modeling the 7-Segment Decoder
// Module BCD_to_7Segment: Modeled using continuous assignment module BCD_to_7Segment (input A,B,C,D, output a,b,c,d,e,f,g); wire T0, T1, T2, T3, T4, T5, T6, T7, T8, T9; assign T0=~A&C; assign T1=~A&B; assign T2=~B&~C; assign T3=~A&D; assign T4=T2&A; assign T5=T2&~D; assign T6=T0&~B; assign T7=T0&~D; assign T8=T1&~C; assign T9=T1&~D; assign a = T0 | T1&D | T4 | T5; assign b = ~A&~B | T2 | ~A&~C&~D | T3&C; assign c = T1 | T2 | T3; assign d = T4 | T5 | T6 | T7 | T8&D; assign e = T5 | T7; assign f = T4 | T5 | T8 | T9; assign g = T4 | T6 | T8 | T9; endmodule assign statements can appear in any order
16
Testing the BCD to 7-Segment Decoder
module Test_BCD_to_7Segment; // No need for Ports reg A, B, C, D; // variable inputs wire a, b, c, d, e, f, g; // wire outputs // Instantiate the module to be tested BCD_to_7Segment BCD_7Seg (A, B, C, D, a, b, c, d, e, f, g); initial begin // initial block A=0; B=0; C=0; D=0; // at t=0 #200 $finish; // at t=200 finish simulation end // end of initial block always #10 D=~D; // invert D every 10 time units always #20 C=~C; // invert C every 20 time units always #40 B=~B; // invert B every 40 time units always #80 A=~A; // invert A every 80 time units endmodule
17
The initial and always Blocks
There are two types of procedural blocks in Verilog The initial block Executes the enclosed statement(s) once The always block Executes the enclosed statement(s) repeatedly until simulation terminates The body of the initial and always blocks is procedural Can enclose one or more procedural statements Procedural statements surrounded by begin β¦ end execute sequentially #delay is used to delay the execution of the procedural statement Procedural blocks can appear in any order inside a module Multiple procedural blocks run in parallel inside the simulator
18
Simulator Waveforms All sixteen input test cases of A, B, C, D are generated between t=0 and t=160ns. Verify that outputs a to g match the truth table.
19
Hierarchical Design Why Hierarchical Design?
To simplify the implementation of a complex circuit What is Hierarchical Design? Decompose a complex circuit into smaller pieces called blocks Decompose each block into even smaller blocks Repeat as necessary until the blocks are small enough Any block not decomposed is called a primitive block The hierarchy is a tree of blocks at different levels The blocks are verified and well-document They are placed in a library for future use
20
Example of Hierarchical Design
Top Level: 16-input odd function: 16 inputs, one output Second Level: Five 4-input odd functions Third Level: Three 2-input XOR gates 4-Input Odd Function z x0 x1 x2 x3 x4 x5 x6 x7 x8 x9 x10 x11 x12 x13 x14 x15 x0 x1 x2 x3 z 16-Input Odd Function x0 x1 x2 x3 x4 x5 x6 x7 x8 x9 x10 x11 x12 x13 x14 x15 z Hierarchical Design typically includes blocks of different functions and size
21
Top-Down versus Bottom-Up Design
A top-down design proceeds from a high-level specification to a more and more detailed design by decomposition and successive refinement A bottom-up design starts with detailed primitive blocks and combines them into larger and more complex functional blocks Design usually proceeds top-down to a known set of building blocks, ranging from complete processors to primitive logic gates
22
Hierarchical Design in Verilog
// Module Odd_4: 4-input Odd function uses three xor gates module Odd_4 (input [0:3] x, output z); wire [0:1] w; xor g1(w[0], x[0], x[1]); xor g2(w[1], x[2], x[3]); xor g3(z, w[0], w[1]); endmodule // Module Odd_16: 16-input Odd function module Odd_16 (input [0:15] x, output z); wire [0:3] w; Odd_4 block0 (x[0:3], w[0]); Odd_4 block1 (x[4:7], w[1]); Odd_4 block2 (x[8:11], w[2]); Odd_4 block3 (x[12:15], w[3]); Odd_4 block4 (w[0:3], z); Five instances of the Odd_4 module
23
Bit Vectors in Verilog A Bit Vector is multi-bit declaration that uses a single name A Bit Vector is specified as a Range [msb:lsb] msb is most-significant bit and lsb is least-significant bit Examples: input [0:15] x; // x is a 16-bit input vector wire [0:3] w; // Bit 0 is most-significant bit reg [7:0] a; // Bit 7 is most-significant bit Bit select: w[1] is bit 1 of vector w Part select: x[8:11] is a 4-bit select of x with range [8:11] The part select range must be consistent with vector declaration
24
Testing Hierarchical Design
Exhaustive testing can be very time consuming (or impossible) For a 16-bit input, there are 216 = 65,536 test cases (combinations) For a 32-bit input, there are 232 = 4,294,967,296 test cases For a 64-bit input, there are 264 = 18,446,744,073,709,551,616 test cases! Testing a hierarchical design requires a different strategy Test each block in the hierarchy separately For smaller blocks, exhaustive testing can be done It is easier to detect errors in smaller blocks before testing complete circuit Test the top-level design by applying selected test inputs Make sure that the test inputs exercise all parts of the circuit
25
Iterative Design Using identical copies of a smaller circuit to build a large circuit Example: Building a 4-bit adder using 4 copies of a full-adder The cell (iterative block) is a full adder Adds 3 bits: ai, bi, ci, Computes: Sum si and Carry-out ci+1 Carry-out of cell i becomes carry-in to cell (i +1) c0 Full Adder a0 b0 s0 c1 a1 b1 s1 c2 a2 b2 s2 c3 a3 b3 s3 c4 ci Full Adder ai bi si ci+1
26
Full Adder Full adder adds 3 bits: a, b, and c Two output bits:
Carry bit: cout Sum bit: sum Sum bit is 1 if the number of 1's in the input is odd (odd function) sum = (a ο
b) ο
c Carry bit is 1 if the number of 1's in the input is 2 or 3 cout = aΒ·b + (a ο
b)Β·c Truth Table a b c cout sum 0 0 0 0 0 1 1 0 1 0 0 1 1 1 0 0 1 0 1 1 1 0 1 1 1
27
Full Adder Module (Gate-Level Description)
module Full_Adder(input a, b, c, output cout, sum); wire w1, w2, w3; and (w1, a, b); xor (w2, a, b); and (w3, w2, c); xor (sum, w2, c); or (cout, w1, w3); endmodule a b c sum cout w1 w2 w3 Full_Adder
28
16-Bit Adder with Array Instantiation
// Input ports: 16-bit a and b, 1-bit cin (carry input) // Output ports: 16-bit sum, 1-bit cout (carry output) module Adder_16 (input [15:0] a, b, input cin, output [15:0] sum, output cout); wire [16:0] c; // carry bits assign c[0] = cin; // carry input assign cout = c[16]; // carry output // Instantiate an array of 16 Full Adders // Each instance [i] is connected to bit select [i] Full_Adder adder [15:0] (a[15:0], b[15:0], c[15:0], c[16:1], sum[15:0]); endmodule Array Instantiation of identical modules by a single statement
29
16-Bit Adder with Continuous Assignment
// Input ports: 16-bit a and b, 1-bit cin (carry input) // Output ports: 16-bit sum, 1-bit cout (carry output) module Adder_16b (input [15:0] a, b, input cin, output [15:0] sum, output cout); wire [16:0] c; // carry bits assign c[0] = cin; // carry input assign cout = c[16]; // carry output // assignment of 16-bit vectors assign sum[15:0] = (a[15:0] ^ b[15:0]) ^ c[15:0]; assign c[16:1] = (a[15:0] & b[15:0]) | (a[15:0] ^ b[15:0]) & c[15:0]; endmodule
30
Testing the 16-bit Adder Exhaustive testing: 216 Γ 216 Γ 2 = 8,589,934,592 test cases Let us choose only: 3 Γ 3 Γ 2 = 18 test cases Input test cases for a[15:0] = 'h158A, 'h52AF, 'hB903 Chosen randomly with values shown in hexadecimal 'h158A (hexadecimal) = 'b0001_0101_1000_1010 (binary) Underscores are ignored (used to enhance readability) Input test cases for b[15:0] = 'h7095, 'h9A4E, 'hC6BD Also chosen randomly with values shown in hexadecimal Radix symbol: 'b (binary), 'o (octal), 'd (decimal), 'h (hex) Input test cases for cin = 0, 1
31
Writing a Test bench for the 16-bit Adder
module Test_Adder_16; // Test bench for Adder_16 reg [15:0] A, B; reg Cin; // Data and Carry inputs wire [15:0] Sum; wire Cout; // Sum and Carry outputs Adder_16 Test (A, B, Cin, Sum, Cout); // Instantiate 16-bit adder initial begin A='h158A; B='h7095; Cin=0; // Initialize A, B, Cin #10 Cin=1; // t=10, Change Cin #10 B='h9A4E; Cin=0; // t=20, Change B and Cin #10 Cin=1; // t=30, Change Cin #10 B='hC6BD; Cin=0; // t=40, Change B and Cin #10 Cin=1; // t=50, Change Cin #10 A='h52AF; B='h7095; Cin=0; // t=60, Change A, B and Cin #10 Cin=1; // t=70, Change Cin #10 B='h9A4E; Cin=0; // t=80, Change B and Cin // more test cases #10 $finish; // End simulation end endmodule
32
Generating Test Cases in parallel with always
module Test_Adder_16; // Test bench for Adder_16 reg [15:0] A, B; reg Cin; // Data and Carry inputs wire [15:0] Sum; wire Cout; // Sum and Carry outputs Adder_16 Test (A, B, Cin, Sum, Cout); // Instantiate 16-bit adder initial begin A='h158A; B='h7095; Cin=0; // Initialize A, B, Cin #200 $finish; // At t=200 end simulation end always begin // Change A every 60 ns #60 A='h52AF; #60 A='hB903; #60 A='h158A; always begin // Change B every 20 ns #20 B='h9A4E; #20 B='hC6BD; #20 B='h7095; always #10 Cin = ~Cin; // Invert Cin every 10 ns endmodule
33
Simulator Waveforms The values of A, B, and Sum are shown in hexadecimal Can change the radix to binary, octal, and decimal The values of Sum and Cout can be verified easily All 18 test cases of A, B, and Cin are generated (t=0 to 180 ns)
Similar presentations
© 2024 SlidePlayer.com. Inc.
All rights reserved.